Country Classification. Opportunities, Risks, Harm and Parental Mediation.

Opportunities, Risks, Harm and Parental Mediation

July 2013

Country Classification

Ellen J. Helsper, Veronika Kalmus, Uwe Hasebrink,

Bence Sagvari and Jos De Haan

with members of the EU Kids Online network

ISSN 2045-256X

www.eukidsonline.net

Country Classification: Opportunities, Risks and Parental Mediation

2

Please cite as: Ellen J. Helsper, Veronika Kalmus, Uwe Hasebrink, Bence Sagvari and Jos De Haan. (2013). Country Classification: Opportunities, Risks, Harm and Parental Mediation LSE, London: EU Kids Online.

Acknowledgements: The authors would also like to thank Lucyna Kirwil, Catherine Blaya and Sonia Livingstone for their contributions in writing this report.

Previous reports and publications from EU Kids Online include:

Livingstone, S., Kirwil, L, Ponte, C. and Staksrud, E., with the EU Kids Online Network (2013) In their own words: What bothers children online? LSE, London: EU Kids Online. http://eprints.lse.ac.uk/48357/

D'Haenens, L, Vandonink, S. and Donoso, V. (2013) How to cope and build resilience. LSE, London: EU Kids Online. http://eprints.lse.ac.uk/48115/

Livingstone, S., Ólafsson, K., O’Neill, B and Donoso, V. (2012) Towards a better internet for children: findings and recommendations from EU Kids Online to inform the CEO coalition. LSE, London: EU Kids Online. http://eprints.lse.ac.uk/44213/

Haddon, L., Livingstone, S. and the EU Kids Online network (2012) EU Kids Online: National perspectives. LSE, London: EU Kids Online. http://eprints.lse.ac.uk/46878/

Smahel, D., Helsper, E., Green, L., Kalmus, V., Blinka, L. & Ólafsson, K. (2012) Excessive internet use among European children. LSE, London: EU Kids Online. http://eprints.lse.ac.uk/47344/

Dürager, A. & Livingstone, S. (2012) How can parents support children's internet safety? http://eprints.lse.ac.uk/id/eprint/42872

Livingstone, S., Haddon, L., Görzig, A., and Ólafsson (2011) EU Kids Online Final Report. http://eprints.lse.ac.uk/39351/

Livingstone, S., Haddon, L., Görzig, A. and Ólafsson, K. (2011) Risks and safety on the internet: The perspective of European children. http://eprints.lse.ac.uk/33731/

Livingstone, S., Haddon, L., Görzig, A., & Ólafsson, K. (2011) Disadvantaged children and online risk. http://eprints.lse.ac.uk/39385/

Livingstone, S., Ólafsson, K. and Staksrud, E. (2011) Social networking, age and privacy. http://eprints.lse.ac.uk/35849/

Sonck, N., Livingstone, S., Kuiper, E. and de Haan, J. (2011) Digital literacy and safety skills. http://eprints.lse.ac.uk/33733/

Hasebrink, U., Görzig, A., Haddon, L., Kalmus, V. and Livingstone, S. (2011) Patterns of risk and safety online. http://eprints.lse.ac.uk/39356/

Görzig, A. (2011) Who bullies and who is bullied online? A study of 9-16 year old internet users in 25 European countries. http://eprints.lse.ac.uk/39601/

O’Neill, B., Livingstone, S. and McLaughlin, S. (2011). Final recommendations for policy, methodology and research. http://eprints.lse.ac.uk/39410/

Livingstone, S. and Ólafsson, K. (2011) Risky communication online. http://eprints.lse.ac.uk/33732/

Livingstone, S., Haddon, L., Görzig, A. and Ólafsson, K. (2011) Risks and safety on the internet: The perspective of European children: Full findings. http://eprints.lse.ac.uk/33731/

The EU Kids Online network has been funded by the EC Safer Internet Programme in three successive phases of work from 2006-14 to enhance knowledge of children’s and parents’ experiences and practices regarding risky and safer use of the internet and new online technologies.

As a major part of its activities, EU Kids Online conducted a face-to-face, in home survey during 2010 of 25,000 9-16 year old internet users and their parents in 25 countries, using a stratified random sample and self-completion methods for sensitive questions. Now including researchers and stakeholders from 33 countries in Europe and beyond, the network continues to analyse and update the evidence base to inform policy.

For all reports, findings and technical survey information, as well as full details of national partners, please visit www.eukidsonline.net

http://eprints.lse.ac.uk/48357/





http://eprints.lse.ac.uk/id/eprint/42872











http://www.eukidsonline.net/

3

CONTENTS

EXECUTIVE SUMMARY 4

INTRODUCTION 7

1. OPPORTUNITIES 11

1.1. INDICATORS ......................................................... 11

1.2. OPPORTUNITIES: GROUPS OF CHILDREN .................... 12

1.3. OPPORTUNITIES: CLASSIFICATION OF COUNTRIES ......... 13

1.4. OPPORTUNITIES: RECOMMENDATIONS ...................... 17

2. RISK AND HARM 19

2.1. INDICATORS ......................................................... 19

2.2. RISK AND HARM: GROUPS OF CHILDREN .................... 20

2.3. RISK AND HARM: CLASSIFICATION OF COUNTRIES ........ 21

2.4. RISK AND HARM: RECOMMENDATIONS ..................... 24

3. PARENTAL MEDIATION 26

3.1. INDICATORS ......................................................... 26

3.2. PARENTAL MEDIATION: GROUPS OF CHILDREN ............ 26

3.3. PARENTAL MEDIATION: CLASSIFICATION OF COUNTRIES 28

3.4. PARENTAL MEDIATION: RECOMMENDATIONS ............. 30

4. CLASSIFICATION OF COUNTRIES BASED ON OPPORTUNITIES, RISK, HARM AND PARENTAL MEDIATION 32

OVERALL SUMMARY AND RECOMMENDATIONS 36

BIBLIOGRAPHY 41

Annex 1: EU Kids Online 42

OVERVIEW ...................................................................... 42

WORK PACKAGES ............................................................. 42

INTERNATIONAL ADVISORY PANEL ....................................... 42

Annex 2: The network 43

COUNTRY ....................................................................... 43

NATIONAL CONTACT INFORMATION ..................................... 43

TEAM MEMBERS ............................................................. 43


4

EXECUTIVE SUMMARY

This report updates and deepens the

understanding of cross-national differences

among the countries surveyed in EU Kids

Online. Where the previous classification was

based simply on the percentage of children in

each country who used the internet daily, and

who had encountered one or more risks, this

report examines the range and type of online

opportunities, risks and harm experienced by

the children in each country. It also takes into

account the ways in which parents mediate or

regulate their children’s internet use in each

country.

Clusters of countries are most clearly

distinguished in terms of sexual content

risks. Children who are bullied or who give

away personal data are uniformly distributed

across the countries. Using these and many

other factors, the report identifies four country

clusters overall: unprotected networkers,

protected by restrictions, semi supported risky

gamers, and supported risky explorers.

This new analysis reveals that differences

within countries are substantially larger than

differences between countries, whether

measured in terms of online opportunities, risk

of harm or forms of parental mediation. The

advantage of such pan-European similarities

is that it makes sense for policy makers in one

country to learn from the best practice initiated

in another.

New classification of online opportunities, risks, harm and parental mediation clusters

5

On the other hand, the analysis also makes it

clear that, to anticipate the online experience

of any individual child, a host of factors must

be considered – merely knowing where they

live is insufficient as a guide to the

opportunities or risks they may experience.

Findings detailed in this report give hope that

parents’ mediation strategies will develop

positively and constructively alongside the use

of their children’s internet use. Nevertheless,

based on the patterns of children’s online

risks, harm and parenting practices across

Europe there is the possibility of a more

negative pattern developing in some

countries.

There is concern that both too much

parental restriction in the protected by

restrictions cluster and the lack of support

for children’s online use in the

unsupported networkers cluster might

lead to higher levels of harm when risk is

encountered.

Supported risky explorers (Denmark,

Finland, the Netherlands, Norway and

Sweden)

This cluster has more children who are

experienced social networkers. They

encounter more sexual risks online and their

more parents are actively involved in guiding

their children’s internet use.

Parental mediation might co-evolve with risk

and opportunity taking by children: as children

gain more experience and encounter more

risks, parents engage more actively in

safeguarding their internet use. There is,

however, a relatively small group of

vulnerable children in these countries that

experience similar levels of risk to their peers

but lack the parental mediation and

opportunities also enjoyed by their peers.

Policy makers should therefore support

parents and schools, and stimulate industry

players to enhance responsible practices in

relation to internet safety, including seeking to

reach and support those few vulnerable

children may ‘get lost’ in an environment full of

experts.

Semi-supported risky gamers (Bulgaria,

Cyprus, Czech Republic, Estonia, Poland and

Romania)

In these countries, children encounter only

moderate online opportunities, mainly focused

on entertainment, and games in particular. Yet

they still experience relatively high levels of

risk and harm: some encounter a specific risk,

others a range of risks.

Parents undertake rather diverse types of

mediation in these countries, including active

and restrictive forms of mediation, although it

seems these are relatively ineffective. This

may be because the online opportunities and

associated digital skills have only emerged

relatively recently in these countries, so

supportive structures and good practice are

not yet established.

Although parents seem to be trying strategies

across the board, further investigation is

needed to understand why levels of risk are

relatively high and what further interventions

would be beneficial to encourage

opportunities and reduce harm.

Protected by restrictions (Belgium, France,

Germany, Greece, Ireland, Italy, Portugal,

Spain, Turkey and the UK)


6

Children’s online experiences in this cluster of

countries is characterised by relatively low

levels of risk probably because internet use is

also more limited and largely restricted to

practical activities. While parents might be

glad that their restrictive mediation practices

prevent risk, it does seem that they may miss

out on many of the online opportunities.

The question for policy makers, parents and

educators in these countries is whether

opportunity uptake can be increased while

simultaneously limiting more extensive risk of

harm. It is possible that this could be achieved

by a move away from more restrictive forms of

mediation towards more active mediation

patterns.

Such an approach would have to

acknowledge that risk will thereby result, and

further investigation is needed to see whether

children can become sufficiently resilient to

cope with risk when they encounter it.

Unprotected networkers (Austria, Hungary,

Lithuania and Slovenia)

Finally, there is a cluster of countries where

children’s experiences are fairly narrow but

potentially problematic: the social aspects of

Web 2.0 seem to have been taken up with

gusto and the children subsequently

encounter risks but not as much harm, from

being in contact with these opportunities.

Here the challenge is that parents are not as

involved in their children’s internet use as in

the supported risky explorers cluster that they

otherwise resemble, probably because, as

with the semi-supported risky gamers, the

internet is a relatively recent addition in many

families, especially for the parents.

7

INTRODUCTION

Cross-national research is done for several

purposes. Some EU Kids Online reports focus

largely on how different factors related to

children’s internet use, risks, harm and

parental mediation vary between young

people in Europe (Dürager & Livingstone,

2012; Hasebrink et al, 2009; Livingstone et al,

2011; Livingstone & Olafsson, 2011; Lobe et

al, 2011). Some have gone beyond that and

have looked at how these relationships vary

between individual countries. Cultural

differences, information and communication

technology (ICT) diffusion and policies, family

dynamics, the educational system and other

country-specific traditions and values have all

been suggested as influential factors for

country differences in internet opportunities

taken up, risks and harms encountered and

parental mediation (Haddon et al, 2012;

Helsper, 2012).

However, one important and slightly different

approach is to not just compare individual

countries but to group them in terms of

similarities and differences. This approach is

taken in this report and allows for a cross-

European view, where countries are not seen

in isolation but as linked to others. By

grouping countries into larger clusters,

stakeholders will be able to learn from

similarities in policies and initiatives across

these countries and, when other clusters of

countries have different, possibly preferred

characteristics, to change according to what

has worked elsewhere. This can be

understood as a benchmarking approach

where countries with largely similar

characteristics but different outcomes are

seen as examples of how positive outcomes

might be achieved or negative outcomes

avoided. Similarly, much can be learned from

countries that have relatively similar outcomes

in terms of opportunities, risk and harm but

that have taken different policy and mediation

strategies to get to that point.

Meaningful and rigorous segmentation means

an increased understanding of the landscape

of children’s internet use and safety in

Europe, but for this segmentation to be

meaningful and useful, the selection of

characteristics that goes into this clustering of

countries is very important. At an earlier

stage, the EU Kids Online project proposed a

simple and easy to understand classification

of countries in terms of higher and lower risk

and higher or lower use.

The original classification (see Figure 1) was

based on just two indicators, the percentage

of young people who used the internet daily

and the number of risks they ran. While easy

to grasp and relatively intuitive, it did not give

such a detailed insight into how children in

different countries were using the internet. It

missed out on other elements that were

measured by EU Kids Online and which are a

key part of the evidence base for

stakeholders: the harm that children suffered

from the risks encountered and the way in

which parents reacted to this.

In this report we analyse the EU Kids Online

data to come to a new classification of

countries which is slightly more complex and


8

paints a richer picture than the older model.

This report uses the data in the survey on the

range of opportunities young people

encounter online, the types of risks that young

people experience online and whether they

subsequently suffer harm from these and, last

but definitely not least, the strategies that

parents employ across Europe to safeguard

their children’s well-being online.

Figure 1: First EU Kids Online country classification for use and risk

Based on Lobe et al (2011). Cross-national comparisons of risks and safety on the internet. www.EUKidsOnline.net

This report is divided into four sections: (1)

clustering of online opportunities; (2)

clustering of risk and harm; (3) clustering of

parental mediation strategies; and (4) overall

classification of countries based on the

opportunities, risk and harm, and mediation

dimensions. The first three sections have the

same structure:

First, we group children based on the

range of opportunities (Section 1), risks

and harm (Section 2) and parental

mediation strategies (Section 3) that they

encounter.

Second, we describe the characteristics of

these groups of children in Europe

including their age, gender and the

educational level of their parents.

Third, we cluster the European countries

based on the distribution of these groups

of children in each of the European

countries.

Fourth, we describe the characteristics of

these clusters of countries.

9

Section 4 uses the distribution of the groups

of children in each of the European countries

across all three indicators (opportunities, risk

and harm, and parental mediation) to come to

an overall classification of European

countries.

Figure 2: New country classification of online opportunities, risk, harm, and parental

mediation

This approach allows us to show how

European children cluster in terms of

opportunity, risk and harm, and parental

mediation characteristics, and how these

clusters or groups of children are distributed

differently over the different countries, leading

to a classification of countries based on the

presence of different types of children and

their home environments. We thus developed

a more sophisticated approach to classifying

countries (see Figure 2) than the higher/lower

use, higher/lower risk classification that we

had before.

Since there was no clear theoretical

framework that helped us predict which

countries were likely to cluster together in

which way across the three dimensions of

opportunities, risk and harm, and mediation,

this short report is by necessity more

descriptive. In an ad hoc fashion we do try to

explain why some countries might differ on

these dimensions and what might explain the

overall clustering when the three dimensions

are clustered together. The report ends with

recommendations for stakeholders based on

the classifications presented in the four

sections.


10

NOTE ON METHODOLOGY

All sections use the following statistical

techniques. First, scales were created in

relation to the topic of the particular section. A

short description is given in each section of

these scales and how they were created.

Then cluster analyses were used to group

children according to the scores on these

scales. Subsequently, the proportion of

children in each group was calculated for the

different countries. These proportions were

used in a cluster analysis in each section that

led to a classification of countries for each

section. The final step (Section 4) used a K-

means cluster analysis on the 25 countries

based on the proportion of children in each

group per country. Ward’s method with

Squared Euclidean Distance was used for this

section’s clustering.

In the report Ns are for the total sample,

weighted by overall weight. Multi-variate

analyses were performed without weights.

(For more detailed information, see the

Methodological Appendix.)

http://www.lse.ac.uk/media@lse/research/EUKidsOnline/EU%20Kids%20III/Classification/Methodological-Appendix-EU-Kids-Online-country-clustering-report.pdf

11

OPPORTUNITIES

Indicators

We start the classification of countries in this

report with a more in-depth look at what lies

behind the earlier established differences in

use (see Figure 1). One main aim of the EU

Kids Online project was to understand how

opportunities and risks relate to each other. To

do this it is important to go beyond quantitative

measures of use or take-up of opportunities

such as frequency or duration, and come to a

definition that incorporates an understanding of

varieties of internet use. In examining cross-

national patterns it is possible to look at which

types of activities are taken up by certain

children in specific countries and which other

types of activities might be more extensively

engaged with by others in other countries. This

approach incorporates an idea of quality as well

as quantity of use.

The indicators for children’s online

opportunities assessed in the EU Kids Online

survey include duration of use. This indicator

provides plausible information on the

quantitative presence of the internet in young

people’s everyday lives. It reflects the temporal

resources young people devote to online

activities and defines the time frame for more or

less opportunities. To an extent, this indicator

reflects young people’s interests and needs.

Those who expect more gratification and more

opportunities from using the internet will spend

more time on it, making these indicators

plausible predictors of the importance of the

internet in everyday life (Livingstone & Helsper,

2010). The breadth of activities that the young

people undertake is regarded as a strong

indicator of patterns of online usage; therefore,

the number of activities that children undertake

(out of 17 measured) was also included in the

analyses presented here.

More was needed for an in-depth

understanding of how countries cluster beyond

these more quantitative indicators of

opportunities (Hasebrink, 2012). Therefore, we

conducted a factor analysis of the 17 activities

mentioned earlier. These were assessed in the

survey through the options ‘never’ (0) to ‘almost

every day’ (4). To the standard 17 indicators for

online activities, we added two aspects of

online behaviour that seemed particularly

important: having an own profile on a social

networking site (0 = no profile, 1 = one profile, 2

= more than one profile) and a range of five

other activities that are related to contacting or

being contacted by others online (0–5 activities

score possible). The factor analysis led to the

following classification of activities.

Factor 1 (‘Communication’): visiting social

networking profiles is the marker variable

and the factor includes several activities

that are mainly communicative. These

activities are closely related to peer-to-peer

communication.

Factor 2 (‘Creativity’): although the loadings

are rather moderate, all activities involve

some degree of creativity or productivity.

Factor 3 (‘Gaming’): this factor clearly

represents gaming and activities linked to it.

Factor 4 (‘Learning’): the main variable is

using the internet for schoolwork, but also

includes reading or watching news on the

internet.


12

Opportunities: Groups of children

Duration of use, range of activities and the

above four factors were subjected to cluster

centre analyses and a solution with six clusters

was selected which groups children clearly into

different types of users.

Restricted learners (N=7,175)

Children in this group are characterised by a

small amount of online use and a small range

of activities. Participation in all activities is

infrequent, with learning activities the most

frequent. This is the youngest of all of the

groups.

Young networkers (N=3,036)

Children in this group have moderately higher

values for all activities with remarkable

exceptions for the learning activities.

Communication and network activities are

particularly popular. There are proportionally

more girls in this group than in the others.

Moderates (N=5,904)

This group undertakes a wider range of

activities than in the young networkers cluster;

some activities are more integrated into the

group’s everyday practices, particularly learning

activities. Communication and network

activities are less often taken up than the other

activities.

All round explorers (N=2,732)

This group spends almost two hours per day on

the internet and has the biggest range of

activities and highest frequency of online

activities. The least popular, creative activities

are the most popular in this group. Boys are

overrepresented in this group.

Intensive gamers (N=2,729)

This group, with proportionally more boys, has

the longest duration of daily online use (three

hours per day) and a smaller range of activities

than those in the all-round explorer group but

still above the overall average. Gaming

activities have the highest values among all the

groups. Learning activities score relatively low

as well as creative activities.

Experienced networkers (N=3,564)

This group uses the internet slightly more

frequently and for a wider range of activities

= girls = boys = older girls = older boys

Each large icon stands for 1,000 children; smaller icons indicate less than 1,000 children

thousand children

13

than the average user. The most obvious

characteristic is an almost complete absence of

gaming activities. Other opportunities are taken

up almost as frequently as in the all-round

explorers group; communication and network

activities are especially popular. There are

proportionally more girls in this group which is

also the oldest on average of the different

groups of children.

The group descriptions reveal two general

findings. On the one hand, they support ‘the

more, the more…’ rule, according to which time

spent online, range of activities and

engagement in more specific activities are

positively correlated. This is the case for the

restricted learners whose lower frequencies of

use coincide with a narrower range of activities

engaged in and for the all-round explorers who

engage frequently and with a large range of

activities. This corresponds with the idea that

there is a ladder of opportunities whereby

activities are added at every step of increased

experience with the internet, and justifies the

approach taken in the first country

classification, whereby the average level of

frequency of use was taken as an indicator of

higher or lower use and used to group

countries. On the other hand, the groups of

children arrived at through cluster analysis

show that patterns of use do not completely

follow this rule. Young networkers and

moderates are the same age and show almost

the same duration of use, but moderates take

up a range of opportunities while young

networkers are more limited in the range of

activities. Intensive gamers spend by far the

longest time online but engage in only a

moderate range of activities.

Opportunities: Classification of countries

The distribution of the six groups of children in

each country was used to create a country

classification based on opportunities. The

proportion of children in each group is shown in

Table 1.

Table 1 shows which countries have the

highest and lowest percentages of children in

certain user groups. Turkey has the highest

percentage of children in the restricted learners

group (45%) and the lowest number in the

experienced networkers group (5%). Sweden

has the lowest percentage of children in the

restricted learners group (9%) and the

moderates user group (8%). In Ireland just

under half of the children fall in the young

networkers group (45%) and only 4% in the

intensive gamers group, the lowest percentage

of children in that group in all countries. France,

on the other hand, has the lowest percentage

of young networkers (7%). Poland leads in

terms of the proportion of children in the

moderates group (38%), positioning it at the

opposite end of the spectrum of Sweden.

Sweden has the largest (17%) and Spain and

Ireland the lowest proportion of children (6%) in

the all-round explorers group. Bulgaria and

Cyprus join Lithuania with high proportions of

children in the intensive gamers group (20%

each), while Norway leads in terms of

experienced networkers (28% of children fall in

that group), while Greece, like Turkey, has 5%

in this group.


14

Table 1: Percentages of children in each group per country

Restricted

learners

Young

networkers Moderates

All round

explorers

Intensive

gamers

Experienced

networkers

Austria 20 28 15 15 7 15

Belgium 19 18 25 15 6 16

Bulgaria 20 16 22 11 20 11

Cyprus 18 13 28 14 20 7

Czech Republic 16 11 28 13 19 12

Denmark 14 19 25 9 19 13

Estonia 14 11 30 12 19 15

Finland 19 19 23 10 11 19

France 34 7 17 15 5 22

Germany 35 12 21 10 8 14

Greece 29 19 27 9 12 5

Hungary 16 28 24 14 10 8

Ireland 24 45 13 6 4 9

Italy 33 10 27 10 8 12

Lithuania 10 16 24 16 20 14

Netherlands 14 20 28 9 11 19

Norway 14 18 11 14 15 28

Poland 20 13 38 7 13 9

Portugal 29 8 36 11 8 8

Romania 26 9 28 11 19 7

Spain 29 14 35 6 5 12

Sweden 9 27 8 17 14 26

Slovenia 12 31 18 11 12 16

Turkey 45 10 23 11 6 5

UK 25 12 20 14 15 15

All countries 22 17 24 12 12 13

Note: Red bold indicates the countries with the lowest proportions of children in the group and blue bold indicates the countries with the

highest proportions of children in the group.

To come to a more informative classification of

the countries based on the data presented in

Table 1, the following procedure was used: for

each country the proportion of children in each

group was calculated, and this data was used

to run a hierarchical cluster analysis with

countries as cases. This procedure identifies

groups of countries that are quite similar to

each other and quite different to the other

groups. This means that there is still

considerable variety within the clusters as

regards the presence of groups of children in

each country.

A five-cluster solution was decided on after

close examination of the data (see Figure 3).

While countries cluster together in terms of the

opportunities groups of children engage with,

15

the path taken by countries to end up in a

particular opportunities cluster will vary and

might be based in different diffusion and policy

histories.

Young networkers cluster (N=4) (Austria,

Hungary, Ireland and Slovenia)

The countries in this cluster have relatively the

largest representation of children in the young

networkers group (33%, ranging from 28% in

Austria and Hungary to 45% in Ireland);

intensive gamers (8%) are underrepresented in

comparison with the other clusters.

Diversity cluster (N=8) (Belgium, Bulgaria,

Czech Republic, Denmark, Estonia, Finland,

Lithuania and the Netherlands)

This cluster of countries equally represents the

whole range of groups identified with only the

restricted learners group underrepresented

(16%). In this cluster the proportion of children

in all other five groups is close to the average

for all countries.

Moderates cluster (N=6) (Cyprus, Greece,

Poland, Portugal, Romania and Spain)

Almost one third (32%) of the children in these

six countries belongs to the moderates user

group; this is above the average for all

countries and higher than in other clusters,

while the more advanced patterns of use – all-

round explorers (10%) and experienced

networkers (8%) – are underrepresented in

comparison to the other clusters.

Restricted learners cluster (N=5) (Germany,

France, Italy, Turkey and the UK)

The dominant group of children in this cluster of

countries are the restricted learners (34%).

Children in this cluster of countries are least

likely to belong to the young networkers (10%)

and the intensive gamers (8%) groups.

Advanced cluster (N=2) (Norway and

Sweden)

This two-country cluster has by far the highest

representation of the more ‘advanced use’

groups of children, the all-round explorers

(15%) and particularly the experienced

networkers (27%). In addition, the young

networkers (22%) group is overrepresented.

Consequently these countries have on average

the lowest percentage of restricted learners

(12%) and moderates (9%).

The new classification presented in Figure 3

adds understanding to the previous one (see

Figure 1) and gives a more detailed image of

the opportunities engaged with by groups of

children in the different countries.

For example, Norway and Sweden are clearly

‘higher use’, falling in this type of category in

both the first and the new classifications. The

new classification in Figure 3 shows that in this

case they are especially made up of children

who are expert in communication and

networking. The children in these countries are

not only intensive users of the internet but also

extensive users.


16

Figure 3: Map of opportunities country clusters

Eastern and North Eastern European countries

that were classified as ‘higher use’ in the first

classification (Czech Republic, Estonia,

Lithuania and Finland) as well as the

Netherlands and Denmark can be distinguished

from these two Scandinavian countries in the

following way: they belong to the diversity

cluster which means that while there are indeed

quite a few children that belong to expert and

all-round user groups, they also have a

reasonable representation of restricted learners

and moderates. In other words, the landscape

of children’s engagement with opportunities is

more varied.

Poland and Romania previously also classified

as ‘higher use’ can now be looked at differently:

while there might be a relatively high frequency

of use, the predominant pattern is actually a

group of children that is moderate in the range

of opportunities that it engages with.

The UK, previously also classified as higher

use, is an interesting case where the group of

children who are restricted learners has a high

representation which takes it out of the group

with purely higher users and qualifies it as part

of a cluster of countries where most children

make only limited use of the wider range of

activities and focus in this more limited use on

practical applications, such as learning and

information seeking.

France, Italy, Germany and Turkey also fall in

the restricted learners cluster that corresponds

more closely to their earlier classification of

lower use. Hungary and Austria previously

classified as lower use fall in the young

17

networkers cluster, suggesting that the lower

use classification in the previous setting was

partly determined by there being more younger

and less experienced users in these countries.

Belgium, previously also classified as lower

use, falls in the diversity cluster in the new

classification. This might indicate that quantity

is not the only thing that should be taken into

account; while the average child in Belgium

might be an infrequent user, this does not

reflect that the country contains children from

all the different types of groups. In other words,

countries in the diversity cluster are made up of

a variety of groups of children, each of which

takes up different opportunities but might, on

average, have individual children whose use is

not broad or frequent.

Finally, Spain and Portugal were also both

classified as lower use in the previous

classification and have, in the new

classification, become part of the group of the

moderates cluster of countries. The low

presence of the group of children with a high

level of engagement places them in both

classifications in a cluster of countries where

use is infrequent and not as broad as in other

clusters of countries.

Opportunities: Recommendations

Internet use is on the rise in all European

countries, both in terms of number of online

participants as well as the amount and diversity

of use. In this environment, a closer inspection

of patterns of use across European countries is

informative to identify where children seem to

be missing out on internet-related opportunities.

The analysis in this section shows that the

patterns of use greatly vary across European

countries, but also that a clustering of countries

based on similar practices is still possible. This

clustering is largely detached from

geographical proximity and illustrates specific

characteristics of children’s internet use within

countries and the resemblance with specific

other countries.

Stakeholders in these countries should

recognise this variety in designing initiatives

around take-up of opportunities for children.

Here we give tentative recommendations for

countries in different clusters; these should be

contextualised within specific local and national

policy and diffusion characteristics.

Diversity cluster (Belgium, Bulgaria, Czech

Republic, Denmark, Estonia, Finland, Lithuania

and the Netherlands)

These countries with a wide range of different

groups of children (i.e., the diversity cluster)

could think of a learning from peers strategy

whereby the more expert and widely engaged

groups of children serve as champions or

buddies for those who are less broadly

engaged.

Restricted learners (Germany, France, Italy,

Turkey and the UK) and moderates clusters

(Cyprus, Greece, Spain, Poland, Portugal and

Romania)

The strategy of peer-to-peer learning might be

less effective in countries where the majority of

children fall into groups that have low,

infrequent or narrow use. In the restricted

learners cluster children engage with some

practical school-related activities but do not

take advantage of the breadth of opportunities

on offer. Here a policy, education or public

awareness-driven intervention might be more

suited since there is less opportunity for


18

learning from peers. These strategies could

consider relatively basic awareness-raising

campaigns around the opportunities that are

already taken up, as well as stimulating interest

for other, more interactive, creative and

reflexive activities than those that the children

are already engaged in.

Young networkers (Austria, Hungary, Ireland

and Slovenia) and advanced clusters (Norway

and Sweden)

These clusters of countries are on the opposite

end of the spectrum from the other clusters in

terms of use; here most children fall into groups

in which there is high engagement with a

narrower range of activities. However, there

might be a similar approach to that

recommended in the restricted learners cluster,

exposing children to the other opportunities that

are available and broadening their engagement

through policy, education or social marketing

campaigns. In these countries, these strategies

should cater to the young people’s greater

experience and probably engage at a higher

level of sophistication since these young people

are already relative experts in specific areas of

use.

19

RISK AND HARM

Indicators

One of the main aims of the EU Kids Online

project was to go beyond looking at the

activities that children engage in online and

adult definitions of those as risky. It was

important to look at the results of children’s

engagement with these types of activities and

to understand whether any harm was caused

by undertaking what might be labelled ‘risky’

activities (Livingstone et al, 2011). Previous

research has shown that opportunities and

risks are clearly related, that those children who

spend more time online and take up a wider

range of online activities are also more likely to

come across risky online situations (Livingstone

& Helsper, 2010). If risks actually turn into

negative experiences or harm, it depends on

individual and social factors such as self-

confidence, skills and mediation (Smahel et al,

2013; Vandoninck et al, 2013). These

differences between children in the number of

risks they encounter and the extent to which

they experience harm is the starting point of the

next step of our analysis that aims at a more

nuanced understanding of how countries can

be classified in terms of their different types of

risk and harm landscapes.

Here we used a number of indicators from the

EU Kids Online questionnaire to go beyond a

simple classification of children as experiencing

less or more risk by distinguishing different

types of risks and including in this an estimate

of the harm that comes from these different

types of risk. Since the data were nested (i.e.,

only children who experienced a risk could

experience the related harm) it was necessary

to create separate scales for each of the three

risky activities that children were asked about:

seeing sexual images, meeting strangers and

bullying. Since sexting, the fourth risk

systematically described in the full report

(Livingstone et al, 2011), was not asked of 9- to

10-year-olds, it was left out of the estimations.

The newly constructed scales ran from 0–6 –

from no experience of the risk to the child

experiencing the risk online and being very

upset. The rest of the scale was divided as

follows: 1 = risk offline but not online (not used

in analysis); 2 = risk online but not bothered; 3

= risk and bothered but not upset; 4 = risk and

a bit upset; 5 = risk and fairly upset. Since the

questions varied across risk the scales, for

bullying ‘2’ marked children who indicated

being bullied but did not respond to the

subsequent question on being upset and ‘3’

marked children who answered that question

and indicated not being upset. For meeting

strangers, ‘1’ indicated that they had made

friends online but did not meet them offline, ‘2’

that they had met them offline but that they

were not bothered, and the rest replicates the

above scale.

A scale that summed the number of risks

related to exposing personal data and

interaction risks was also included. This contact

risk scale ranged from 0–5 risks encountered.

No follow-up questions about harm were asked

for this scale.

A two-step cluster analysis procedure

appropriate for combining scale and interval

variables was used to group the children. This


20

included only children who had experienced at

least one of the three main risks online (i.e.,

scores 2–6 on those scales). This approach

created a de facto cluster of children who had

not encountered any risk online (N=19,420).

Across all countries 5,722 children had

experienced one or more risks online and 519

encountered all the risks. When we included

this cluster of those who did not experience any

risk, just two groups of children appeared:

those who had not experienced any risks and

those who had (very few children had

experienced harm). While this is a reflection of

a low risk and harm reality, it would lead to a

similar classification to the one presented in

Figure 1 which does not allow for clustering of

countries based on the different types of risks

and harm experienced by children; we

therefore only clustered those children who had

encountered at least one risk online.

Risk and Harm: groups of children

When using the indicators described earlier, a

cluster analysis revealed a solution identifying

three groups of children that fit the data. The

three individual risk scales were more

influential in determining the clusters than the

contact risk scale. Overall experiences of risk

and especially harm were low and thus higher

risk and harm in this classification is relative,

that is, in comparison to the average and other

groups. The three groups can be described as

follows:

Sexual risk and harm group (N=2,299)

These children experience relatively high

sexual images risks and higher levels of harm

for this risk. While they experience bullying and

have met people offline, they experience only

moderate levels of harm (compared to children

in the higher risk/harm and contact risk

clusters). They also have the lowest score on

the contact risk scale. This group of children do

not stand out in terms of their age but boys are

more likely to fall into this group and their

parents are more likely to have tertiary

education than those in any of the other

groups.

Higher risk /harm group (N=1,250)

These children experience relatively higher

levels of risk across all risk categories except

the overall contact risk scale. They are

especially more likely to experience higher

levels of harm from online bullying and meeting

strangers offline. This group of children does

not stand out in terms of their age but girls are

more likely to belong to this than to any other

group, and their parents are more likely to have

secondary education than those of other

groups.


Each large icon stands for 1,000 children; smaller icons indicate less than 1,000 children

21

Contact risks group (N=2,172)

These older children (average age = 15) are

most likely to experience harm from meeting

people offline and score highest on the overall

contact risk scale, related to giving out personal

information. Boys are more likely to fall in this

group than girls and the distribution of parental

education follows an average pattern.

No online risk group (N=19,420)

This group consists of younger children

(average age = 12 years) who did not

encounter any risk online and have thus not

experienced any harm. Girls and boys are just

as likely to be part of this group and, in

comparison to other groups, parental education

levels are lower.

The grouping of children shows that there is no

clear linear trend from no risk and harm to

higher risk and harm. Children who are likely to

encounter one particular type of risk are not

necessarily more likely to encounter other types

of risk and harm. While there was one group of

children (i.e., the higher risk/harm group) who

experienced risks across the board, there were

also two distinct groups of children who

experienced one type of risk and harm but were

not exposed to risk or harm from other types of

activities. For example, older children in the

contact risk group seem to avoid other types of

risks and have overall lower levels of harm. The

children in the sexual risks group, while not

managing to avoid other risks, experience little

harm from other online risks.

Risk and Harm: Classification of countries

The distribution of the children in each group

was used to create a country classification

based on risk and harm. Table 2 shows what

proportion of children in each country falls

within the different groups identified in the

previous section.

Table 2 shows that Italy had the highest

percentage of children in the no risk group of

children (90%), while Estonia had the lowest

percentage (59%) of children in that group.

Norway had the highest (20%) and Germany

the lowest percentage (4%) of children in the

sexual risks group. Romania had the highest

percentage (12%) of children in the higher

risk/harm group while Portugal and Italy had

the lowest percentage (2%) of children in that

group. Lithuania had the highest percentage

(22%) of children in the contact risks group,

placing it at the opposite end of the spectrum

from Italy and Turkey, which had only 4% of

children in that group.

To come to a more informative classification of

countries based on the distribution of children

over these groups, a cluster analysis was

conducted following a two-step clustering

procedure that allows for the inclusion of

categorical and interval variables. The no risk

group was left out of the analysis to make the

classification more informative for describing

the patterns of risk and harm across countries.

This procedure identifies groups of countries

that are quite similar to each other and quite

different to the other groups. This means that


22

there is still considerable variety within the

clusters as regards the presence of groups of

children in each country. Thus, while countries

cluster together in terms of the risks and harm

groups of children encounter, the historical

routes that lead to this particular risk and harm

classification will vary for different countries

within the same cluster.

Table 1: Percentages of children in each group per country

No risk Sexual risks Higher risk and harm Contact risks

Austria 72 9 6 13

Belgium 71 12 5 11

Bulgaria 73 11 5 11

Cyprus 80 10 4 6

Czech Republic 65 13 6 16

Denmark 65 16 10 10

Estonia 59 13 10 18

Finland 68 17 4 11

France 70 13 6 11

Germany 79 4 4 13

Greece 81 8 3 8

Hungary 80 9 5 7

Ireland 83 8 4 5

Italy 90 5 2 4

Lithuania 63 12 4 22

Netherlands 72 18 3 6

Norway 61 20 6 12

Poland 78 10 4 8

Portugal 81 9 2 8

Romania 67 8 12 13

Spain 81 9 4 6

Sweden 64 11 9 16

Slovenia 67 14 4 14

Turkey 84 9 3 4

UK 79 7 8 6

EU Average 73 11 5 10



The three clusters of countries can be

described as follows (see Figure 4):

Higher risk/harm cluster (N=10) (Austria,

Belgium, Bulgaria, Czech Republic, Estonia,

France, Lithuania, Romania, Slovenia, and

Sweden)

In this cluster there are relatively high

percentages of children in each of the risk and

23

harm groups. They have the highest proportion

of children in the higher risk/harm group (7%)

and in the contact risks group (15%); 12% of

children in these countries fall into the sexual

risks group, which is second only to the sexual

risks cluster of countries.

Lower risk/harm cluster (N=11) (Cyprus,

Germany, Greece, Hungary, Ireland, Italy,

Poland, Portugal, Spain, Turkey and the UK)

In this cluster of countries there is the lowest

proportion of children in each of the risk and

harm clusters: 8% of children fall in the sexual

risks group, 4% in the higher risk/harm group

and 7% in the contact risks group.

Sexual risks cluster (N=4) (Denmark, Finland,

the Netherlands and Norway)

The countries in this cluster are mainly

characterised by a high proportion of children in

the sexual risks group (18%); the proportion of

children in the higher risk/harm group is the

second highest (6%) among all the countries,

and the proportion of children in the contact

risks group is relatively low (10%).

Figure 4: Map of risk and harm country clusters

Figure 4 shows a slightly different classification

for the countries based on risk and harm than

the first country classification presented at the

beginning of this report. A distinction can be

made especially among those countries that

were considered as higher risk and some risk in

the first classification (see Figure 1).

In the previous classification, Denmark and

Norway were classified as higher risk. In the

new classification, Norway and Denmark are


24

part of a small cluster of countries that has

most children in the sexual risks group and has

thus been labelled the sexual risks cluster. The

Netherlands and Finland, previously classified

as some risk, make up the other two countries

of this small cluster of four countries where the

largest proportion of children fall into the sexual

risks group.

In the new classification, the remaining

Scandinavian country, Sweden (previously

classified as higher risk), falls into a relatively

large cluster of countries labelled the higher

risk/harm cluster. The new higher risk/harm

classification in Figure 4 offers additional

information in comparison to the higher risk

classification presented in Figure 1.

In countries in the higher risk/harm cluster the

largest number of children does indeed fall in

the higher risk/harm group with relatively varied

risks and associated harm. In addition, these

countries have a high proportion of children that

run mostly contact risks. While there are fewer

children in the sexual risks group than in the

countries of the sexual risks cluster, the

proportion of children in that group is still

relatively high. The Eastern European

countries, Estonia, Lithuania, the Czech

Republic, Bulgaria and Romania, also fall in the

higher risk/harm cluster of countries in the new

classification. Poland, and Slovenia, previously

classified as some risk (see Figure 1), are now

also part of this higher risk/harm cluster.

Most interestingly Belgium, France and Austria,

previously classified as lower risk, fall in the

higher risk/harm cluster under the new

classification (see Figure 4). This can be

explained if a single child in this country does

not experience a great number of risks, but

many children in these countries belong to a

group of children that experiences relatively

great harm from bullying or runs contact or

sexual risks. In other words, it is probably not

quantity or breadth but quality or type of risk

encountered that makes the difference here.

The other countries that were previously

classified as lower or some risk remain in a

similar category in the new classification. The

UK, Turkey, Spain, Portugal and Poland were

all classified as some risk and fall in the cluster

of lower risk countries in the new classification.

While there are still some children that fall into

the sexual, contact and higher risk/harm

groups, these are relatively low in comparison

to the other clusters of countries. Ireland,

Germany, Italy, Hungary and Greece were all

classified as lower risk in the previous

classification and remained in the lower

risk/harm cluster in the new classification.

Risk and Harm: Recommendations

Countries looking to learn from the situation in

other countries would be wise to pay attention

not only to the quantity but also the quality of

the risks that different groups of children

encounter in these countries.

There are qualitative differences between

countries in terms of the range and type of risks

that children experience which lead to a

different classification of countries than when

only looking at quantity. Therefore, when

aiming to build up a strategy to counter risk,

taking the countries that are most suitable for

comparison are probably not those whose

children encounter the same quantity of risks

but rather similar types of risks and harm. This

leads to the following tentative

25

recommendations for specific clusters of

countries, which should be contextualised

within specific local and national policy and

diffusion characteristics.

Higher risk/harm cluster (Austria, Belgium,

Bulgaria, France, Czech Republic, Estonia,

Lithuania, Romania, Slovenia and Sweden)

This is a large cluster of higher risk and harm

countries in which children fall into a variety of

different groups that encounter specific risks. In

this cluster of countries, campaigns could be

built for these different groups of children as a

one size fits all strategy would likely be

ineffective in reaching all of them. Messages

about generalised risk and harm might reach

children who are part of the higher risk/harm

group in these countries, but might seem

irrelevant to the also relatively large proportion

of children who encounter mostly sexual risks.

An understanding of which children are most

likely to fall into which risk/harm groups, based

on an understanding of the national context, is

essential to be able to offer these differentiated,

targeted interventions and engage parents and

children.

Sexual risks cluster (Denmark, Finland, the

Netherlands and Norway)

It is not necessarily the case that the groups of

children that run one risk are also more likely to

run other risks. In this cluster of countries, most

children belong to a group that experiences

mainly sexual risks and are unlikely to

encounter other risks or harm. For these

countries, single-issue campaigns focusing on

sexual risks are possibly more effective than

general awareness-raising campaigns against

more general internet risks. In these countries

in particular, general campaigns might scare

most children, who in general avoid broader

risks or, on the contrary, risk children ignoring

these campaigns because they seem irrelevant

to them.

Lower risk/harm cluster (Cyprus, Germany,

Greece, Hungary, Ireland, Italy, Poland,

Portugal, Spain, Turkey and the UK)

While the average child in these countries

might fall into a group that experiences lower

risk or harm, this does not mean that there are

no children who encounter risk or harm in this

cluster of countries. In these countries in

particular, vulnerable children are likely to be

isolated since their experiences differ from that

of the majority. In these countries, schools,

parents, counsellors and organisations that

interact with vulnerable children are likely to be

the best route for support. Individual adults and

peers are likely to be more effective here than

general campaigns since they will be more

aware of the child’s individual circumstances

and can act to prevent the risks and online

opportunities these children encounter leading

to harm for the young person. Linking these

findings to the previous section it should be

noted that in these countries children also tend

to be less broadly engaged with the

opportunities the digital world offers, which

indicates that there is a negative flip side to

lower risks.


26

PARENTAL MEDIATION

Indicators

Looking at the opportunities and risks that

children encounter online is obviously useful to

understand how the situation in different

countries can be compared. However, this has

to be contextualised within the children’s social

environment, and one way of doing this is to

look at whether parents react differently to their

children’s internet use across countries.

Parental mediation has always been

considered an important factor in relation to

children’s media use, and there is a well-

developed field of thinking about the different

types of parenting and how they might be

related to different types of use (Livingstone &

Helsper, 2008). Several papers have been

written based on the EU Kids Online data about

parental mediation and its relation to children’s

internet use, especially risk taking of children

(e.g. Dürager & Livingstone, 2012; Kalmus et

al, in press; Kirwil, 2009; Livingstone &

Haddon, 2008; Paus-Hasebrink et al, 2012;

Sonck et al, 2012).

This section looks in more detail at how these

different types of parenting are distributed

across Europe. A distinctive feature of the EU

Kids Online survey is that it asked a number of

questions about several types of parental

mediation. Furthermore, matched questions

were asked of the child (CQ) and the parent

(PQ) most involved in the child’s internet use.

Three compound scales of parental mediation

were created, based on nine original mediation

scales asked of all children:

Active mediation (ranging from 0–22). The

items on this scale enquired about active

mediation of internet use (CQ + PQ) and

active mediation of internet safety (CQ +

PQ).

Restrictive mediation (ranging from 0–12).

The items in this scale looked at restrictive

mediation practices (CQ + PQ).

Monitoring and technical restrictions

(ranging from 0–12). The items on this

scale measured parental monitoring (CQ +

PQ) and technical restrictions (PQ) put in

place by parents.

Households where monitoring or technical

restrictions were not possible (due to the child

not using the internet at home) were assigned

the value of zero on the monitoring and

technical restrictions scales. For those who had

answered ‘Don’t know’, the average values

were used to replace missing data.

Parental mediation: Groups of children

While many parents are rather passive in

mediating their children’s online behaviour,

others are involved in their online participation

and mediate actively. These parents differ to a

great extent in how often and in what ways they

interact with their children. A cluster analysis of

children examined how children grouped

together based on their parents’ mediation

styles, and found four groups of children.

The four groups can be labelled as follows,

where parental mediation levels are relative,

that is, higher or lower in comparison to other

groups of children:

27

All rounders (N=5,583)

Parents of this group of relatively young

children (average age = 12) practise all three

types of mediation above the overall sample

average; the levels of monitoring and technical

restrictions and active mediation are particularly

high. The parents in this group are the most

likely to have secondary levels of education

and the proportion of parents with primary

education only is the lowest in this group. In

this, like in most other groups, there are a

similar number of boys and girls.

Active mediation preferred (N=7,320)

Parents of the relatively older children (average

age = 14) in this group prefer using active

mediation (though to a somewhat lesser extent

than all-rounders), while practising the other

two types of mediation, especially restrictive

mediation, less than the sample average. The

parents of the children in this group are more

likely than those in the other groups to have

tertiary levels of education.

Restrictive mediation preferred (N=6,350)

Parents of these younger children (average age

= 11) clearly prefer setting rules and restrictions

to the child’s internet use. They also engage in

active mediation, although less than all-

rounders and parents preferring active

mediation, and are clearly less in favour of

monitoring and technical restrictions. The

parents of the children in this group are also

relatively more likely to have primary education

or less.

Passive (N=5,889)

Parents of this group of older children (average

age = 14) practise all three types of mediation

below the overall sample average; the levels of

active mediation and monitoring and technical

restrictions are particularly low. The children in

this group are more likely to be boys than girls

and their parents are most likely to have

primary education.

Only the parents of all-rounders apply a broad

range of mediation techniques, and in the

passive group parents tend to apply a narrow

range of mediation techniques, if any. The

qualitative differences are apparent in the other

two larger groups where parents tend to

prioritise one mediation technique over all

others. This is particularly true for the restrictive

mediation group. Those in the active mediation

group reject the types of mediation most

popular in the restrictive mediation group,

positioning children in these groups on opposite

sides of the spectrum. The difference between


Each large icon stands for 1000 children, smaller icons indicate less than 1000 children


28

these two groups is underlined by differences in

age and the parental educational levels of

children in these groups; children whose

parents are active mediators are older and their

parents tend to be higher educated, while the

reverse is true for children whose parents are

restrictive mediators.

Parental mediation: Classification of countries

These groups of children with parents that use

different mediation types are not equally

distributed across different countries. Table 3

shows how the mediation patterns vary across

Europe.

Table 3: Percentage of children in each of the parental mediation groups per country

All-rounders Active mediation

Restrictive

mediation Passive

Austria 18 20 31 31

Belgium 25 27 25 24

Bulgaria 19 28 23 30

Cyprus 23 37 16 24

Czech Republic 24 39 17 20

Denmark 13 52 12 23

Estonia 21 39 9 31

Finland 19 44 22 15

France 26 21 35 18

Germany 20 15 46 18

Greece 24 19 29 29

Hungary 18 26 25 31

Ireland 30 18 40 12

Italy 27 19 33 21

Lithuania 7 39 9 46

Netherlands 26 45 16 13

Norway 23 51 17 8

Poland 34 34 9 23

Portugal 22 25 26 27

Romania 17 27 23 32

Spain 26 25 28 22

Sweden 22 44 17 17

Slovenia 11 30 18 41

Turkey 16 15 38 30

UK 33 24 26 17

EU average 22 31 24 24



29

Table 3 shows that Poland has the largest

percentage of children (34%) whose parents

apply a wide range of different mediation

strategies and the lowest percentage (9%) of

children in the group where parents apply

mostly restrictive mediation. Lithuania, on the

other hand, has the lowest percentage of

children in the all-rounders mediation group

(7%) but similar to Poland has the lowest

percentage (9%) of children in the restrictive

mediation group. Germany has the lowest

percentage (15%) of children in the group

where parents use active mediation and the

highest (46%) in the restrictive mediation

group. Turkey is also lowest (15%) in the active

mediation group and higher than average

(38%) in the restrictive mediation group.

Denmark, in contrast, is high (52%) in active

mediation and low on children (12%) in the

restrictive mediation group. Estonia joins

Lithuania and Poland in having the smallest

proportion of children (9%) in the restrictive

mediation group. Lithuania leads the way in

terms of passive mediation, with just under half

of the children (46%) falling in that group. This

positions it opposite to Norway, where only 8%

of children are part of the group where parents

mediate passively.

The distribution of the four groups of children

was used to create a country classification

based on parental mediation through a cluster

analysis. While countries cluster together in

terms of the parental mediation styles groups of

children encounter, the routes that lead to this

particular classification based on parents’

mediation practices will vary for different

countries within the same cluster. This analysis

classified countries into four clusters (see

Figure 5):

Figure 5: Map of parental mediation country clusters


30

Restrictive mediation cluster (N=11) (Austria,

Belgium, France, Germany, Greece, Ireland,

Italy, Portugal, Spain, Turkey and the UK)

In this cluster there is the highest proportion of

children in the group whose parents prefer

restrictive mediation (32%). The percentage of

all-rounders (24%) is also above the average in

this cluster of countries, while the proportion of

children in the group whose parents prefer

active mediation is the lowest (21%).

Passive cluster (N=5) (Bulgaria, Hungary,

Lithuania, Romania and Slovenia)

In this cluster of countries there is the highest

proportion of children in the group with passive

parents (36%). The percentages of children in

other groups are below average.

All-rounders cluster (N=4) (Cyprus, the Czech

Republic, Estonia and Poland)

This is the most heterogeneous cluster of

countries with the highest percentage of all-

rounders (25%) and the second highest

proportion of children in the group whose

parents prefer active mediation (37%).

However, the percentage of children in the

group with passive parents is also above the

average in this cluster (25%).

Active mediation cluster (N=5) (Denmark,

Finland, Netherlands, Norway and Sweden)

This cluster of countries is characterised by the

highest proportion of children in the group

whose parents prefer active mediation (47%)

and below the average percentages of children

in all other clusters.

The largest cluster in this classification of

countries based on parental mediation of

children’s internet use is that of restrictive

mediation (see Figure 5). Almost all Western,

Central and Southern European countries fall

into this category. The exception is the

Netherlands, which falls in the active mediation

cluster with the Scandinavian countries. In most

Eastern European countries passive mediation

is more common which positions them in a

different cluster from the Scandinavian

countries. However, the Czech Republic,

Poland and Estonia have a stronger presence

of groups of children whose parents use a

range of mediation strategies and have thus

been labelled the all-rounders. Cyprus also falls

within this cluster while its neighbours, Greece

and Turkey, are part of the biggest cluster of

countries in which the largest number of

children is part of the restrictive mediation

group.

Parental Mediation: Recommendations

From other research we know that the

differences and similarities in parental

mediation between countries are at least partly

based on diffusion rates, parents’ online

experience and technological opportunities,

which in turn are related to national wealth.

Furthermore, there are cultural and social

differences between countries resulting in

differences in parental values and preferred

styles of parenting as well as differences in

welfare state institutions which regulate female

labour force participation and the availability of

public childcare facilities (e.g. Kalmus &

Roosalu, 2012; Kirwil et al, 2009).

31

Policy at a national level can contribute to

internet safety by building on existing practices.

Yet, we need to bear in mind that what is

needed in single countries depends on local

contexts. This leads to the following

recommendations for clusters of countries,

which should be contextualised within specific

local and national policy and diffusion

characteristics:

Passive cluster (Bulgaria, Hungary, Lithuania,

Romania and Slovenia)

The majority of children in these five countries

have parents who are mostly passive in their

mediation. Since these are countries with

relatively low diffusion of the internet, these

parents might lack either awareness or skills to

properly mediate the internet use of their

children. E-inclusion strategies to improve

online participation and digital skills among

parents are appropriate in these countries.

Supporting parents with information on, for

example, how and when to talk to their children

and how to build relationships of mutual trust

might give parents more confidence in their

mediation strategies.

Restrictive mediation cluster (Austria,

Belgium, Germany, Greece, Spain, France,

Ireland, Italy, Portugal, Turkey and the UK)

As was the case for opportunities and risks,

there are clusters of countries where mediation

strategies are relatively homogeneous. In this

homogeneous cluster of countries, the majority

of children have parents that prefer to either not

mediate or to mediate in a restrictive fashion

rather than in an active way. These strategies

have been considered less effective than

others in safeguarding children, and in these

countries a single-issue campaign on

promoting more active mediation with very

specific advice on how this could be done could

be appropriate. It is interesting that varied local

policies and diffusion histories are linked to a

similarity in the way parents mediate their

children’s use across these countries.

All-rounders cluster (Cyprus, the Czech

Republic, Estonia and Poland)

In this cluster of countries, children come from

a wide variety of different groups with parents

using a whole range of different mediation

strategies. In this cluster there is a great

heterogeneity among children’s parents.

Parents combine, for example, active mediation

and restrictive mediation, which means that

they might not be sufficiently aware of what

each kind of mediation means in terms of the

efficiency of their parental effort and, as a

consequence, the online safety of their

children. Campaigns and initiatives targeted at

less knowledgeable parents could therefore be

effective. Since there are quite a few children

with actively mediating parents, peer learning

or parental discussions might be a way forward.

Active mediation cluster (Denmark, Finland,

Netherlands, Norway and Sweden)

Active mediation is seen as the ideal by many

stakeholders, and this cluster of countries has a

majority of children with parents who employ

this strategy. However, even in these countries

there are considerable numbers of children that

have other parental mediation types, and in this

case emphasising the best practices

undertaken by most parents could help the

other parents understand how to help their

children.


32

CLASSIFICATION OF COUNTRIES BASED ON OPPORTUNITIES, RISK, HARM AND PARENTAL MEDIATION

The final step in this report is to bring all the

previous classifications together to come to an

overall informative classification of countries

that includes information on online

opportunities, risks, harm and parental

mediation in the European countries. To do this

we conducted a hierarchical cluster analysis on

the 25 countries using the percentage of

children in the different opportunities, risk and

harm, and mediation groups in each country

(see Tables 1, 2 and 3) as the basis for

analysis. Each of the resulting clusters

describes a combination of the take-up of

internet opportunities, risks and harms

encountered and parental mediation strategies.

This analysis looks at the distribution of groups

of children within each country and which

combinations of these groups are most likely to

be found together in different countries. The

resulting combinations seem to indicate that

parents mediate more actively when children

are more advanced users with a wider range of

risks and opportunities involved. The following

labels were attached to these clusters:

Unprotected networkers: network

opportunities and risks and passive

mediation

Protected by restrictions: practical, few

opportunities and risks and all-round,

restrictive mediation

Semi-supported risky gamers: moderates

and intensive gamers, higher risk/harm and

all-rounders/active mediation

Supported risky explorers: experienced

networkers and sexual risks groups with

active mediation

The proportion of children in each

opportunities, risk and harm, and parental

mediation groups are distributed as follows

over the four different clusters of countries (for

a detailed description of the presence of

clusters per country, see the ‘Individual country

clustering’ reports).

Table 4 shows that there is considerable

variation in the ways in which the groups of

children coincide in the four country clusters.

The following, more detailed, description of the

country clusters shows that the way in which

this distribution of groups of children makes

countries cluster together is not always as

expected (see also Figure 6). It is important to

remember that while countries within each

cluster have similar distributions of groups of

children with specific opportunities, risk and

harm and parental mediation styles, the routes

that lead to membership of a particular cluster

are likely to have varied for countries within that

cluster.

http://www.lse.ac.uk/media@lse/research/EUKidsOnline/EU%20Kids%20III/Classification/Home.aspx

http://www.lse.ac.uk/media@lse/research/EUKidsOnline/EU%20Kids%20III/Classification/Home.aspx

33

Table 4: Percentage of children in each opportunities, risk and harm, and mediation group

per country cluster

Clusters of countries→

Groups of children ↓

Unprotected networkers

Protected by

restrictions

Semi-supported

risky gamers

Supported risky

explorers

European Average

Opportunities

Restricted learners 14 30 19 14 22%

Young networkers 26 16 12 20 17%

Moderate users 20 24 29 19 24%

All round explorers 14 11 11 12 12%

Intensive gamers 12 8 18 14 12%

Experienced

networkers

13 12 10 21 13%

Risk and Harm

No risk 70 80 70 66 73%

Sexual risks 11 8 11 16 11%

Higher risk/harm 5 4 7 6 5%

Contact risks 14 8 12 11 10%

Mediation

All-round 13 25 23 21 22%

Active 29 21 34 47 31%

Restrictive 21 33 16 17 24%

Passive 37 22 27 15 24%

Number of countries 4 10 6 5 25

Note: Red bold indicates the countries with the highest proportions of children in the group and blue bold indicates the countries with the

lowest proportions of children in the group.

Unprotected networkers (N=4) (Austria,

Hungary, Lithuania and Slovenia)

This cluster stands out for having the highest

percentage of young networkers and all-round

explorers, while restricted learners are

underrepresented compared to the average.

The risk pattern of the countries in this cluster

is close to the average, with a slightly higher

number of children in the contact risks group.

This cluster has the lowest percentage of

children in the all-round mediation group (13%),

while most belong to the passive mediation

group (37%). Although Lithuania is

geographically further away, Austria, Hungary

and Slovenia form a special Central European

cluster with children from groups that can be

described as high on network opportunities and

related risks while having parents with more

passive mediation strategies.

Protected by restrictions (N=10) (Belgium,

France, Germany, Greece, Ireland, Italy,

Portugal, Spain, Turkey and the UK)

With 10 countries, this cluster is undoubtedly

the largest one; it not only has the greatest

number of countries, but it also includes the

largest European countries. While in terms of

opportunities, risks and harm and mediation

these clusters show high similarity, their policy

and internet diffusion history is quite disparate.

Therefore, it is important to keep in mind that

similar outcomes can be achieved through


34

different routes. In this report we are not able to

bring in evidence of how different national

policies or values are differentially linked to

similar outcomes. However, it is clear that there

is a pay-off taking place where restriction leads

to relatively lower risk taking but also to a

narrower range of activities as undertaken by

children. This cluster has the highest proportion

of children in the restricted learners group and

very few children in the intensive gamers group

compared to the other clusters. This cluster

further stands out for the few risks that are run:

80% of children also belong to the no risk

group, and the percentage of children in the

other risks groups is the lowest across the

board. Compared to the average levels of all

European countries most children are from the

group where mediation is more restrictive than

active. In geographic terms this cluster

embraces the countries of Western and

Southern Europe.

Semi-supported risky gamers (N=6)

(Bulgaria, Cyprus, Czech Republic, Estonia,

Poland and Romania)

This cluster of countries is the most difficult to

pin down. Their common characteristics are the

higher proportion of children in the moderates

group of opportunities, but they also have a

relatively high proportion of intensive gamers

and a below-average level of young

networkers. They have the highest proportion

of children in the higher risk/harm group (7%)

and average proportions of children in the other

risk and harm groups. In this cluster, the group

of children whose parents apply restrictive

mediation is least frequent compared to other

countries, and while children with actively

mediating parents are most common, the

proportion of children in this group is not higher

than average. Other forms of mediation also

stay around the average. Countries included in

this cluster are mainly from Central and South

East Europe.

Supported risky explorers (N=5) (Denmark,

Finland, the Netherlands, Norway and Sweden)

This cluster is very clearly defined. In this

cluster, most of the children can be found in the

experienced networkers group, while the

representation of the groups of restricted

learners and moderates is well below average.

This intensive level of engagement with

opportunities in networked environments by

older children is accompanied by a relatively

higher proportion of children in groups that

encounter risks, especially sexual risks. Only

66% of children belong to the no risk group,

which is the lowest proportion compared to the

other three clusters, and 16% of children are in

the sexual risks group. There are relatively

fewer children here with parents in the

restrictive mediation group (17%) and a large

proportion of children in the group where

parents prefer active mediation (47%). This

cluster of countries belongs geographically and

culturally to the Scandinavian region which

includes the Netherlands as in previous

separate classifications of opportunities, risk

and harm, and mediation.

35

Figure 6: New country classification based on opportunities, risks and harm, and parental

mediation


36

OVERALL SUMMARY AND RECOMMENDATIONS

This report updates and deepens the

understanding of cross-national differences

among the countries surveyed in EU Kids

Online. Where the previous classification was

based simply on the percentage of children in

each country who used the internet daily, and

who had encountered one or more risks, this

report examines the range and type of online

opportunities and risk of harm experienced by

children in each country. It also takes into

account the ways in which parents mediate or

regulate their children’s internet use in each

country.

The classification of countries presented in this

report shows that while more opportunities are

often found in combination with more online

risks, this understanding needs to be

contextualised for a better understanding of

practices of safeguarding children’s internet

use in European countries. First of all, it should

be noted that for many of the groups of children

with opportunities, risk and harm, and

mediation experiences, differences within

countries are larger than differences between

countries. For example, higher educated

parents are more active in mediating their

children’s internet use than parents with less

education. This implies that policy makers in

single countries should also assess the factors

within their country that contribute to

differences in opportunities, risk and harm, and

parental mediation as well as looking at how to

benchmark their practices against countries in

a similar situation.

In other words, the type of mediation is not the

only factor related to risks or harm at a country

level. Within a country, parental mediation

should be considered in combination with other

influences on and characteristics of young

people such as the role that schools and peers

play, child development and resilience and the

socio-demographic characteristics of their

parents.

It is important to note that, at the country level:

Risks are not necessarily cumulative.

Countries in which a large group of children

encounters a specific risk do not

necessarily also have large groups of

children who encounter other risks or a

greater number of risks.

Opportunities and parental mediation

practices are also not cumulative. If there

is a large group of children who participate

intensively in a particular activity or whose

use is mediated actively, other groups of

children who are intense users in a different

way or whose parents use other passive or

restrictive mediation styles are not

necessarily present in the country.

There is no consistent link between a

particular style of parental mediation

and lower risk and harm and more

opportunities. While there is a cluster of

countries in which restrictive practices co-

occur with lower levels of risk and harm,

countries with the highest proportion of

restrictive mediation practices are not

systematically lower risk and harm

countries.

While it is clear that in countries with a

larger representation of experienced,

intense user groups of children more

specific risks are run, it is not clear that this

37

is linked to greater vulnerability. Instead,

children in these countries seem to be

more supported in their use.

Clusters of countries distinguish themselves

more from others based on their patterns of

content risks, sexual content risks in particular,

than in relation to the contact related risks.

Children who are bullied and run risks by giving

away personal data are more uniformly

distributed across countries when parental

mediation and broader engagement are taken

into account.

The general classification of countries using

opportunities, risk and harm, and parental

mediation as indicators leads us to the

following recommendations for the different

country clusters:

Supported risky explorers (Denmark,

Finland, the Netherlands, Norway and Sweden)

The cluster of Scandinavian countries,

including the Netherlands, in which children

who encounter sexual risks are more strongly

represented is also the cluster of countries

where more experienced networkers can be

found. In these countries, the level of internet

diffusion is also relatively high, with parents

generally more digitally skilled and aware of

online risks compared to other countries. Thus,

these highly experienced risk takers live in an

environment where most of their parents are

actively involved in guiding their use or at least

being there to support them if they need help.

Both parents and children in these countries

are more likely to be (pro)active in their use,

risk taking and mediation. In these countries,

the focus seems to be on supporting children to

develop in a digital environment where risks will

be encountered.

This is an indication that parental mediation

might co-evolve with risk and opportunity

taking by children – as children get more

experienced and encounter more risks,

parents engage more actively in

safeguarding their internet use. It would thus

be erroneous to conclude that, in those

countries where this type of risk taking is

particularly prevalent, active mediation stops

risk taking completely.

The need for concern by parents and policy

makers in this cluster of countries should be

directed towards a relatively small group of

vulnerable children that encounters the same

risks as their peers but is not embedded in the

mediation and opportunity structures that

are common in their country. These children

might still be harmed, and specific care remains

necessary for the few children in need of

protection in this cluster of countries. Policy

makers could support parents and schools and

stimulate industry players to continue their

responsible practices in relation to internet

safety and design targeted strategies to reach

the relatively few vulnerable children who may

‘get lost’ in an environment full of experts.

Semi-supported risky gamers (Bulgaria,

Cyprus, Czech Republic, Estonia, Poland and

Romania)

Perhaps more problematic is the cluster of

countries where most children engage only

moderately or focus on entertainment-related

activities but where still high levels of risks and

subsequent harm can be found. Most children

in these countries fall into groups that

encounter either specific risks or a range of

risks and subsequent harm. Very diverse types

of mediation are practised in these countries

Risks and safety on the internet: The perspective of European children

38

and active mediation is also included but

apparently less effective. Tentatively we might

say that there is not a clear message or policy

in these countries about effective mediation,

and that this is linked to varied risk-taking

patterns. Parents seem to be trying strategies

across the board, but there is probably less

dialogue possible between parents of different

groups of children.

Even though active mediation is also relatively

frequent in these countries, it does not seem to

have the same effect as in the countries where

larger percentages of children are experienced

in internet use. Perhaps this is the result of a

relatively recent take-up of the online

opportunities, and further crystallisation of

interaction processes between parents and

children is still needed as a base for more

internet safety.

For policy makers, raising awareness among

parents on different internet risks and

suitable ways to deal with these might be a

way forward. As indicated in previous sections

of this report, policy makers and other

stakeholders might use the diversity in their

country to start discussions that bring parents

and schools together. Diversity and

heterogeneity offer great opportunities to

discuss different strategies and how these

might work within the particular cultural and

technological landscape that they inhabit.

The configuration of relationships between

mediation, opportunities and risks in the

supported risky explorers and the semi-

supported risky gamers country clusters

suggests that the development of parental

mediation and children’s use practices is a

symbiotic and continuous process. This

development can support children to encounter

some risks that will help them build resilience

without this leading to harm.

Protected by restrictions (Belgium, France,

Germany, Greece, Ireland, Italy, Portugal,

Spain, Turkey and the UK)

Another combination was found in these

countries that have relatively low levels of risk

and harm which are also the countries in which

use is more limited and restricted to practical

activities. While parents might be happy that

their restrictive mediation practices take

children away from higher risk and harm and

sexual or contact risks, it does seem that they

may miss out on many of the online

opportunities. In these countries, the

emphasis seems to have been on safeguarding

children by trying to minimise risk which is

linked to restricting their broader engagement.

The question for policy makers, parents and

educators in these countries is whether

opportunity uptake can be increased while

simultaneously limiting more extensive risk

taking and, even more importantly, harm. It is

likely, from other research, that this means a

move away from more restrictive forms of

mediation towards more active mediation

patterns such as those found in the Nordic

countries.

There might be no such thing as a completely

risk-free environment. Parents who actively

mediate do so not to prevent all risk but to

make sure that children can encounter risks

important for their development and

resilience in an environment that

safeguards them from serious harm. This

could particularly be taken up by industry,

39

which might help in creating technological

platforms that allow for this relatively safe risk

taking and at the same time help children

develop their digital literacy and broaden their

engagement with the internet.

Unprotected networkers (Austria, Hungary,

Lithuania and Slovenia)

Finally, there is a cluster of countries where

children are one-sided in terms of both the risks

and the opportunities that they encounter

online. In these countries the social aspects of

Web 2.0 seem to have been taken up with

gusto, and the children subsequently

experience risks but not as much harm from

being in contact with these opportunities.

Here the issue is that parents are not as

involved in their children’s internet use as in

the cluster of countries where sexual risks and

experienced networkers can be found. This

could mean that, as more children move into

the more intensive all-round user groups,

they might also encounter more risks and

subsequent harm. It is difficult, of course, to

predict how this will play out – countries might

have different diffusion and developmental

trajectories, and thus we do not know whether

such a development would also lead parents to

adapt their mediation strategies.

In summary

This report will help us to understand the way in

which European countries can be classified in

terms of opportunities, risks and harm, and

parental mediation styles that children

experience. The routes that lead to a country

ending up in a specific cluster might have been

very different for countries within that cluster, to

be able to offer differentiated, targeted

interventions and to engage parents and

children, an understanding of the national

context is essential. There are lessons to be

learned from the characteristics of countries in

other clusters as well as from best practices in

countries within the same cluster.

Findings detailed in this report give hope that

parents’ mediation strategies will develop

positively and constructively alongside the use

of their children’s internet use. Nevertheless,

there is a risk of a more negative pattern

developing in the protected by restrictions and

unsupported networkers clusters that limits

children’s engagement or could lead to higher

levels of harm.

Stakeholders concerned about child

development and online safety should not rest

on their laurels and assume that it will all

naturally work out in the end.

In countries with relatively restricted or

moderate levels of use (and risk taking), policy

makers, industry and third sector stakeholders

should work alongside educators and parents

to make sure that further and broader

engagement in the future is not accompanied

by parental mediation strategies that are

passive and likely to lead to more harm, or

restrictive, which is likely to deny children the

opportunities that are available online. Best

practices in clusters of countries with more

experienced users and parents with more

involved mediation strategies can serve as

guidelines for how to achieve this.

Each child will also require individually tailored

mediation related to their social environment

and experiences and, of course, in those rare

cases where a child has come to more serious


40

harm, counselling and trained advisers must be

available.

However, general awareness-raising

campaigns and policies aimed at creating an

online environment where children can take up

opportunities and encounter risks safely are

important to support parents and educators in

their efforts.

41

BIBLIOGRAPHY

Dürager, A. & Livingstone, S. (2012) How can parents support children’s internet safety?, London: EU Kids Online. Retrieved from


Haddon, L., Livingstone, S. & EU Kids Online Network (2012) EU Kids Online: National perspectives, London: LSE, EU Kids Online.

Hasebrink, U. (2012) ‘Young Europeans’ online environments: a typology of user practices’, jn S. Livingstone & A. Goerzig (eds) Children,

risk and safety on the internet, Bristol: The Policy Press.

Hasebrink, U., Livingstone, S., Haddon, L. & Ólafsson, K. (2009) Comparing children’s online opportunities and risks across Europe:

Cross-national comparisons for EU Kids Online, London: LSE, EU Kids Online.

Helsper, E.J. (2012) ‘Which children are fully online?’, in S. Livingstone & A. Goerzig (eds) Children, risk and safety on the internet,

Bristol: The Policy Press.

Kalmus, V. & Roosalu, T. (2011) ‘Parental mediation of EU kids’ internet use revisited: looking for a complex model of cross-national

differences’, International Journal of Media and Cultural Politics, 7, 55–66.

Kalmus, V., Blinka, L. & Ólafsson, K. (in press) ‘Does it matter what mama says: Evaluating the role of parental mediation in European

adolescents’ excessive internet use’, Children & Society.

Kirwil, L. (2009) ‘Parental mediation of children’s internet use in different European countries’, Journal of Children and Media, 3, 394–409.

Kirwil, L., Garmendia, M., Garitaonandia, C. & Fernandez, G.M. (2009) ‘Parental mediation’, in S. Livingstone & L. Haddon (eds) Kids

Online. Opportunities and risks for children (pp. 199–216), Bristol: The Policy Press.

Livingstone, S. & Haddon, L. (2008) ‘Risky experiences for children online: Charting European research on children and the internet’,

Children & Society, 22, 314–323.

Livingstone, S. & Helsper, E.J. (2008) ‘Parental mediation and children’s internet use’, Journal of Broadcasting & Electronic Media, 52(4),

581–599.

Livingstone, S. & Helsper, E.J (2010) ‘Balancing opportunities and risks in teenagers’ use of the internet: The role of online skills and

family context’, New Media & Society, 12(2), 309–329.

Livingstone, S. & Ólafsson, K. (2011) ‘Risky communication online’, in K. Ólafsson, S. Livingstone & L. Haddon (2013) Children’s use of

online technologies in Europe: A review of the European evidence database, London: LSE, EU Kids Online.

Livingstone, S., Haddon, L., Görzig, A. & Ólafsson, K. (2011) Risks and safety on the internet: The perspective of European children. Full

findings, London: LSE, EU Kids Online.

Lobe, B., Livingstone, S., Ólafsson, K. & Vodeb, H. (2011) Cross-national comparisons of risks and safety on the internet, London: LSE,

EU Kids Online.

Paus-Hasebrink, I., Bauwens, J., Dürager, A.E. & Ponte, C. (2012) ‘Exploring types of parent–child relationship and internet use across

Europe’, Journal of Children and Media (JOCAM), Special Issue.

Smahel, D., Helsper, E., Green, L., Kalmus, V., Blinka, L. & Ólafsson, K. (2012) Excessive internet use among European children,

London: LSE, EU Kids Online.

Sonck, N., Nikken, P. & de Haan, J. (2012) ‘Determinants of internet mediation: a comparison of the reports by parents and children’,

Journal of Children and Media (JOCAM), Special Issue.

Vandoninck, S., d’Haenens, L. & Roe, K. (2013) ‘Online risks: Coping strategies of less resilient children and teenagers across Europe’,

Journal of Children and Media, 7(1), 60–78.


42

ANNEX 1: EU KIDS ONLINE

Overview

In its first phase (2006-9), as a thematic network of

21 countries, EU Kids Online identified and critically

evaluated the findings of nearly 400 research

studies, drawing substantive, methodological and

policy-relevant conclusions. In its second phase

(2009-11), as a knowledge enhancement project

across 25 countries, the network surveyed children

and parents to produce original, rigorous data on

their internet use, risk experiences and safety

mediation. In its third phase (2011-14), the EU Kids

Online network is examining findings and critical

analyses of internet and mobile technology uses

and associated risks among children across Europe,

drawing on these to sustain an active dialogue with

stakeholders about priority areas of concern for child

online safety.

Thus, the network has widened its work by including

all member states and extending its engagement –

both proactively and responsively - with policy

stakeholders and internet safety initiatives. It has

also deepened its work through targeted hypothesis

testing of the pan-European dataset, focused on

strengthening insights into the risk environment and

strategies of safety mediation, by pilot testing

innovative research methodologies for the nature,

meaning and consequences of children’s online risk

experiences, and conducting longitudinal

comparisons of findings where available over time.

Last, it is updating its work on the online database of

available findings, and by producing timely updates

on the latest knowledge about new and emerging

issues (for example, social networking, mobile

platforms, privacy, personal data protection, safety

and awareness-raising practices in schools, digital

literacy and citizenship, geo-location services, and

so forth).

Work packages

WP1: Project management and evaluation.

WP2: European evidence base

WP3: Hypotheses and comparisons

WP4: Exploring children's understanding of risk

WP5: Dissemination of project results

WP6: Policy implications

International Advisory Panel

María José Cantarino, Telefonica, Spain.

Michael Dreier, Clinic for Behavioural Addictions

Mainz, Germany.

David Finkelhor. Crimes against Children

Research Center, University of New Hampshire,

USA.

Lelia Green, ARC Centre of Excellence for

Creative Industries and Innovation, Australia.

Natasha Jackson, FOSI and GSMA, UK.

Amanda Lenhart, Pew Internet & American Life

Project, USA.

Janice Richardson, European Schoolnet, and

Insafe, Brussels, Belgium.

Kuno Sørensen, Save the Children, Denmark.

Janis Wolak, Crimes against Children Research

Center, University of New Hampshire, USA.

43

ANNEX 2: THE NETWORK

Country National Contact Information Team Members

AT

Austria

Ingrid Paus-Hasebrink [email protected] Department of Audiovisual Communication, University of Salzburg, Rudolfskai 42, A-5020 Salzburg, Austria

Ingrid Paus-Hasebrink Andrea Dürager Philip Sinner Fabian Prochazka

BE

Belgium

Leen D'Haenens [email protected] Centrum voor Mediacultuur en Communicatietechnologie (OE), OE Centr. Mediacult.& Comm.technologie, Parkstraat 45 – bus 3603, 3000 Leuven, Belgium

Leen d'Haenens Verónica Donoso Sofie Vandoninck Joke Bauwens Katia Segers

BG

Bulgaria

Luiza Shahbazyan [email protected] Applied Research and Communications Fund, 1113, Sofia, 5, Alexander Zhendov St.

Luiza Shahbazyan Jivka Marinova Diana Boteva

HR

Croatia

Dunja Potočnik [email protected] Institute for Social Research, Zagreb

Ivana Ćosić Pregrad Marija Lugarić Dejan Vinković Dragana Matešković

CY

Cyprus

Yiannis Laouris [email protected] Cyprus Neuroscience & Technology Institute Science Unit of the Future Worlds Center 5 Promitheos, 1065 Lefkosia, Cyprus

Yiannis Laouris Elena Aristodemou Aliki Economidou Tao Papaioannou

CZ

Czech Republic

David Šmahel [email protected] Faculty of Social Studies, Masaryk University Joštova 10, 602 00 Brno, Czech Republic

David Šmahel Štepán Konečný Lukáš Blinka Anna Ševčíkov Petra Vondráčková Alena Černá Hana Macháèková Věra Kontríková Lenka Dědková

DK

Denmark

Gitte Stald [email protected] IT University of Copenhagen, Ruud Langgaards Vej 7, 2300 Copenhagen, Denmark

Gitte Stald Heidi Jørgensen

EE

Estonia

Veronika Kalmus [email protected] Institute of Journalism and Communication, University of Tartu, 18 Ülikooli St., 50090 Tartu, Estonia

Veronika Kalmus Pille Pruulmann-Vengerfeldt Maria Murumaa-Mengel Andra Siibak Kersti Karu Lennart Komp Inga Kald Marianne Võime Kairi Talves

FI

Finland

Reijo Kupiainen [email protected] Department of Journalism and Mass Communication, University of Tampere, 33014 Finland

Reijo Kupiainen Kaarina Nikunen Annikka Suoninen Sirkku Kotilainen

FR

France

Catherine Blaya [email protected] IREDU - Université de Bourgogne

Catherine Blaya Elodie Kredens Seraphin Alava Said Jmel

mailto:[email protected]










44

DE

Germany

Uwe Hasebrink [email protected] Hans Bredow Institute for Media Research Warburgstr. 8-10, D - 20354 Hamburg, Germany

Uwe Hasebrink Claudia Lampert

EL

Greece

Liza Tsaliki [email protected] Department of Mass Media and Communications National and Kapodistrian University of Athens 5 Stadiou Street, Athens 105 62, Greece

Liza Tsaliki Despina Chronaki Maria Philippi Sonia Kontogiani

HU

Hungary

Bence Ságvári [email protected] Information Society and Network Research Center – ITHAKA, Perc u. 8, Budapest, 1036 Hungary

Bence Ságvári Anna Galácz

IS

Iceland

Kjartan Ólafsson University of Akureyri Borgum v/Nordurslod, IS-600 Akureyri, Iceland

Kjartan Ólafsson Thorbjorn Broddason Gudberg K. Jonsson

IE

Ireland

Brian O’Neill [email protected] College of Arts and Tourism, Dublin Institute of Technology, Rathmines Road, Dublin 6, Ireland

Brian O’Neill Thuy Dinh Simon Grehan Nóirín Hayes Sharon McLaughlin

IT

Italy

Giovanna Mascheroni [email protected] OssCom, Università Cattolica del S. Cuore Largo Gemelli, 1, 20123 Milano, Italy

Piermarco Aroldi Giovanna Mascheroni Maria Francesca Murru Barbara Scifo

LV

Latvia

Inta Brikše [email protected] Department of Communication Studies University of Latvia

Inta Brikše Skaidrite Lasmane Marita Zitmane Ilze Šulmane Olga Proskurova-Timofejeva Ingus Bērziņš Aleksis Jarockis Guna Spurava Līva Brice Ilze Bērziņa

LT

Lithuania

Alfredas Laurinavičius [email protected] Department of Psychology, Mykolas Romeris University, Ateities st. 20, LT-08303 Vilnius, Lithuania

Alfredas Laurinavičius Renata Mackoniene Laura Ustinavičiūtė

LU

Luxembourg

Georges Steffgen [email protected] Université du Luxembourg

Georges Steffgen André Melzer Andreia Costa

MT

Malta

Mary Anne Lauri [email protected] University of Malta

Mary Anne Lauri Joseph Borg Lorleen Farrugia Bernard Agius

NL

Netherlands

Nathalie Sonck [email protected] SCP, Parnassusplein 5, 2511 VX Den Haag, Netherlands

Nathalie Sonck Jos de Haan Marjolijn Antheunis Susanne Baumgartner Simone van der Hof Els Kuiper Natascha Notten Marc Verboord Peter Nikken










45

NO

Norway

Elisabeth Staksrud [email protected] Dept. of Media and Communication, University of Oslo Boks 1093 Blindern, 0317 Oslo, Norway

Elisabeth Staksrud Jørgen Kirksæther Birgit Hertzberg Kaare Ingunn Hagen Thomas Wold

PL

Poland

Lucyna Kirwil [email protected] Department of Psychology Warsaw School of Social Sciences and Humanities ul. Chodakowska 19/31, 03-815 Warsaw, Poland

Lucyna Kirwil Aldona Zdrodowska

PT

Portugal

Cristina Ponte [email protected] Departamento de Ciências da Comunicação Faculdade de Ciências Sociais e Humanas, Universidade Nova de Lisboa (UNL) Av. de Berna, 26-C, 1069-061 Lisboa, Portugal

Cristina Ponte José Alberto Simões Daniel Cardoso Ana Jorge Rosa Martins

RO

Romania

Monica Barbovschi [email protected] Babes-Bolyai University, Faculty of Sociology and Social Work, 21 Decembrie 1989 st. no.128-130, Cluj-Napoca, Romania

Monica Barbovschi Eva Laszlo Bianca Fizesan Gyöngyvér Tőkés George Roman Valentina Marinescu Anca Velicu

RU

Russia

Galina Soldatova [email protected] Moscow State University, Foundation for Internet Development

Galina Soldatova Ekaterina Zotova Elena Rasskazova Polina Roggendorf Maria Lebesheva

SK

Slovakia

Jarmila Tomková [email protected] VUDPaP, Institute for Child Psychology and Pathopsychology

Jarmila Tomková Ľudmila Václavová Magda Petrjánošová Dana Petranova Norbert Vrabec Magdalena Petrjanosova

SI

Slovenia

Bojana Lobe [email protected] Centre for Methodology and Informatics Faculty of Social Sciences, University of Ljubljana Kardeljeva pl. 5, Ljubljana, Slovenia

Bojana Lobe Sandra Muha

ES

Spain

Maialen Garmendia [email protected] Depto. de Sociología, Universidad del País Vasco, Apartado 644, 48.080 Bilbao, Spain

Carmelo Garitaonandia Maialen Garmendia Gemma Martínez Miguel Angel Casado Estefanía Jiménez

SE

Sweden

Cecilia von Feilitzen [email protected] The International Clearinghouse on Children, Youth and Media, Nordicom, Goteborg University, Box 713, 405 30 Goteborg, Sweden

Cecilia von Feilitzen Elza Dunkels Olle Findahl Ulrika Sjöberg Karl Dahlstrand

CH

Switzerland

Sara Signer [email protected] IPMZ - Institute of Mass Communication and Media Research, Andreasstrasse 15, CH-8050 Zürich

Sara Signer Martin Hermida Heinz Bonfadelli

TR

Turkey

Kursat Cagiltay [email protected] Department of Computer Education and Instructional Technology, Faculty of Education, Middle East Technical University, 06531, Ankara, Turkey

Kursat Cagiltay Engin Kursun Turkan Karakus Secil Tisoglu












46

UK

United Kingdom

Coordinator, Management Group

Leslie Haddon [email protected] Department of Media and Communications London School of Economics and Political Science Houghton Street, London WC2A 2AE, UK

Sonia Livingstone Leslie Haddon Cornelia Reyes Ellen Helsper John Carr


Country Classification. Opportunities, Risks, Harm and Parental Mediation.

Documents

Transcript of Country Classification. Opportunities, Risks, Harm and Parental Mediation.